Dummettian Anti-Realism

I find it interesting that realists almost unfailingly present discussions of realism in terms of truth, when, by all anti-realist accounts (and by that I mean Dummett, Wright, and Tennant), meaning is the central notion. I think this is disingenuous, and a violation of that great principle of academic writing, that one is to represent one's opponents' views in the best possible light.

While I don't, at present, have at hand any good passages or soundbites from Dummett or Wright to emphasize this point, the point is easily found in Tennant (1987), being the first paragraph of chapter one:

Anti-Realism is a doctrine about language, meaning, and logic. It rests on four powerful ideas:

1. when we are mastering language, and when we are exercising that mastery, all that is available to us for gleaning or conveying meanings is the overt, observable behaviour of fellow speakers;
2. the meaning of any well-formed expression in our language depends in a rule-governed way on the meanings of its constituent simple expressions;
3. when we have mastered a language, its sentences have a determinate meaning for us;
4. our knowledge of those meanings can be displayed by an appropriate exercise of recognitional capacities shared by competent speakers.

Principle (1) we have from the later Wittgenstein. Principle (2) we have had ever since the early Frege. Principle (3) we reaffirm pace Quine. Principle (4) is due to Dummett. It will assume special significance in the course of our investigations.

Now, given such a clear statement, how is it fair for the realist to present anti-realism as Christopher Norris (2002) does here (from the second paragraph of chapter one):

Dummett is known chiefly for his anti-realist approach to issues of meaning and truth, an approach that grew out of his intensive study of the work of Gottlob Frege. In this account the meaning of a statement is given by its truth-conditions, which in turn derive from our ability to manifest a knowledge of the kinds of situation to which it properly applies. that is to say, our only criterion of truth or falsehood is the capacity we have to verify or falsify the statement under review. So when Frege holds that 'sense determines reference' Dummett takes this to entail that it is simply unintelligible to posit the existence of objective ('verification-transcendent') truths that somehow surpass or exceed the limits of our best-attainable knowledge. In which case we cannot--or should not--make claims about truth or reality beyond whatever can be borne out by our proof-procedures, information sources, or practices of reasoning on the evidence to hand.

I fail to see how expressing the views in this manner amounts to anything like charity of exposition. The issue is not that these things are false (though I think the case could be made that some of them are), it's that the emphasis is inappropriate. Consider the emphasis in the following sentence from Norris' introduction (p. 7):

[If the anti-realist is right], we should then have to conclude that mathematicians working on a solution to Goldbach's Conjecture quite literally have no idea of what might count as an adequate formalised proof, or that astronomers quite literally have no conception of the state of affairs (i.e., the existence or non-existence of a duplicate solar system) that would serve to decide the truth-value of any statement concerning it.

As before, the sentence is true, but the emphasis is unfair. This is not the way the sentence is true that is supposed to be persuasive. Instead, the emphasis should be as follows:

[If the anti-realist is right], we should then have to conclude that mathematicians working on a solution to Goldbach's Conjecture quite literally have no idea of what might count as an adequate formalised proof, or that astronomers quite literally have no conception of the state of affairs (i.e., the existence or non-existence of a duplicate solar system) that would serve to decide the truth-value of any statement concerning it.

The anti-realist doesn't argue that the mathematicians or astronomers have nothing in mind, but rather only that it is indeterminate what it is that is in mind, that there is no one conception that the mathematicians or astronomers necessarily share, and that this indeterminacy can't be clarified by experience or by an improvement to the language.

There are still manners of expressing what the mathematicians and astronomers might have in mind, such as illustrations, metaphors, or analogies, but these aren't the kinds of things that are appropriate bearers of truth, seeing as how they aren't literal indicative sentences of a language. (The possibility that a different conception of knowledge, say knowledge-how, might be useful in this situation is something I've pondered, but since this conception is not the target of the anti-realist, the point is a red herring.)

I think the realist argument can't be taken seriously until one of them responds to the challenge to present an adequate account of meaning. I'll leave you with Dummett's words (1978, p. 423) on one such attempt:

Kripke's theory admits a distinction, absent from Frege, the Tractatus and the positivists alike, between epistemic and metaphysical necessity. From my point of view, Kripke's account is worse than the Tractatus/positivist one, in that, not only does it make meaning correlative to necessity, but it overtly chooses metaphysical rather than epistemic necessity as the type of necessity to which it is correlative. This represents an extreme form of repudiation of thesis (i) [meaning has to do with knowledge]. But I cannot see that a repudiation so extreme can relate to a notion that can be recognised as one of meaning. It has the effect that Kripke's notion of 'meaning' can have only the use of serving to give a semantics for sentences involving metaphysical necessity: the notion had therefore better be re-labelled 'intension', and we now no longer have any theory of meaning.